Vol. 5, No. 3, March 1966
NOTES 479
This is most likely due to the fact that X is too large to use only second-order perturbation theory. For symmetry D d d i t can be shown that gll = 2.0023 cos 2a:
gl = 2.0023 cos2 a:
2
4
+ 2k sin2 a
+ d g k sin 2 a
sin2 a:
15 +4 14
6 sin 2 a ]
(5)
1-5q
2
rl=
X H(xz, yz) - E(3z2 - r 2 )
TABLE 11
VALUES FOR k ,
P ,A N D KPFOR Mu6+AND W f k
Ion
~(10-4 %~(10-4 cm-1) cm-1)
Mo6+in TiOp -44.0 -33.0 Mo6+in (NHa)sInCls~HzOb -48.2 -44.0 Mo5f in CaWOIC Mog5 0.32 -26.0 -24.6 &/Ios7 0.32 -26.9 -25.4 Mo(CN)s30 . 0 8 -24.3 -28.6 w(cN)s30.07 38.5 42.9 a See ref 14 and 15. See ref 16. G. H. Azarbayejani and A. L. Merlo, Phys. Rev., A137,489 (1965).
ion. The smaller magnitude of KP could also mean more delocalization but it could also be due to a small amount of hybridization of 5s with the 4d3,2--,2 orbital since this is allowed in D d d symmetry. Such hybridization makes a negative contribution to K and hence would reduce the magnitude of KP. Similar comparisons are difficult to make for wj since less data are available on the esr of W5+ compounds. Garifyanov, et a1.,18have reported that A = 146 X cm-' for WO(SCN)52-, and if it assumed that K is similar to that in Mo5+compounds then P 90 x cm-1, which is considerably larger than that found for W(CN)g3-. The fact that 811 < gl for M O ( C N ) ~and ~ - W(CN)g3when present in K I M o ( C N ) ~and K4W(CN)8 suggests that d,, is the ground state. This is consistent with the duodecahedral structure that would be expected in the crystal state since Mo(CN)g4- in K J I O ( C N ) ~ ~ ~ H ~ O possesses this structure.2 It has been f0und~912~ that gll < gl for Cr5f in CrOs3-, which is known to have the duodecahedral structure. The fact that (9) for these ions in the duodecahedral configuration is much different from that of the ions in solution is additional evidence that these ions do not possess the duodecahedral configuration in water solution. f
where an attempt to take into account covalent binding is made by allowing k to be less than 1 (its value for a pure d orbital). Values of P,k, and KP are given in Table I1 along with values obtained for Mo5+ in other environments. For Mo6+ in Ti02 and (NH4)21nC15.H20 eq 3 was used and eq 5 was used for Mo5f in CaWOJ since this appears to be another case where d322-72is the ground state. In the calculations it was assumed that A and B were both positive for Mog5vg7 since g N is negative for these isotopes. The rather small values of k indicated the inadequacy of trying to represent the effects of covalency by the method employed in eq 5 . Since P is proportional to (r-3)av the smaller value for M O ( C N ) ~ must ~ - mean that the unpaired electron is more delocalized, hence more covalent in bonding orbitals, than it is for Mo6+ in Ti02 and (NH4)2InCl5* H20. The small value for Caw04 is most likely the result of a $5 ion occupying the site of the smaller +6
-
(18) N. S. Garifyanov, B. M. Kozyrev, and V. N. Fedotov, Dokl. A k a d . Nauk S S S R , 166, 641 (1964). (19) B. R. McGarvey, J . Chem. Phys., 31, 2001 (1962). (20) J. D. Swalen and J. A. Ibers, ibid., 3 1 , 17 (1962).
Notes CONTRIBUTION FROM THE DEPARTMENT OF CHEMISTRY, action of the metals with halogens (Clz, Br2,12) in acetoCASEINSTITUTE OF TECHNOLOGY, CLEVELAND, OHIO 44106 r ~ i t r i l e . ~ -The ~ majority of the chloride and bromide
Preparations and Properties of Chromium(I1) Complexes. 1V.I Complexes with Acetonitrile and Observations on Tetrahedral Chromium(I1) BY DAVIDG. HOLAHAND JOHN P. FACKLER, JR. Received August 23, 1965
Acetonitrile forms a variety of complexes with halides of first-row transition metals,2 either by recrystallization of the halide from acetonitrile or by re-
complexes of Mn(II), Fe(II), Co(II), Ni(II), and Zn(I1) are of the type MX2.2CH3CN. Apart from the Co(I1) and Zn(I1) complexes, they have octahedral structures with (presumably) bridging halides. The iodides of Mn(II), Fe(II), Co(II), and Ni(I1) form complexes of formula MI2*3CH&!N which have been s h ~ w n ~to- ~ contain octahedral and tetrahedral metal ions, as in (1) Part 111: D. G. Holah and J. P. Fackler, Jr., Inoug. Chem., 4, 1721 (1965). (2) R. A. Walton, Q ~ ~ a t 'Rev. l . (London), 19, 126 (1965). (3) D. G. Holah, Ph.D. Thesis, University of Hull, 1963. (4) B. J. Hathaway and D. G. Holah, J. Chem. Soc., 2400 (1964). (5) B. J. Hathaway and D. G. Holah, ibid., 2408 (1964). (6) B. J. Hathaway and D. G. Holah, ibid., 537 (1965)
480
NOTES
[M(CH3CN)6][MI~].The reactions of Ti, V, and Cr metals with halogens in acetonitrile give, for the most part, simple octahedral addition compounds3sfiof the type MX3.xCH3CN. The most stable complexes of copper(I1) halides and acetonitrile appeara to be CuX2.CH3CN(X = C1, Br), although7 CuC12.2CH3CN and8 CuC12.2/3CH3CNhave been reported. The crystal structures* of CuC12. xCH3CN (x = 1 and 2/3) show the presence of dimeric or polymeric units with halogen bridges and nitrogencoordinated CH3CN groups. Continuing our studies' of 3d4systems, chromium(I1) halide complexes with acetonitrile have been prepared. It was thought that these complexes might be particularly interesting in view of the tetrahedral species formed with the iodides of Mn(I1) through Ni(I1). Experimental Section All preparations and measurements were carried out under inert atmospheric conditions as previously described.9 Dichlorobis(acetonitrile)chromium(II) .-Chromium( 11) chloride, anhydrous or hydrated,Qis dissolved in the minimum quantity of hot ethanol, to which hot acetonitrile is added until a solid begins to crystallize. On cooling, pale blue crystals appear which are filtered off, washed with acetonitrile, and dried a t room temperature under vacuum. Anal. Calcdfor CrCI2.2CH3CN:C,23.43; H , 2.95; Cr, 25.37; C1,34.59. Found: C, 23.2; H,3.1; Cr,25.4; C1, 34.6. Blue crystals, probably the complex contaminated with ZnClp 2CH3CN,3 can be obtained by reducing a solution of chromium(111) bromide in acetonitriles with Zn-Hg. Dibromobis(acetonitrile)chromium(II).-Hexaaquochromium(11) bromideg is dehydrated by heating to 150" under high vacuum for 24 hr and transferred to a Soxhlet extraction thimble. The halide is extracted under nitrogen with dry acetonitrile (distilled from P 2 0 5 ) . Slowly, pale blue-green needles form in the extraction flask. The crystals, which are oxidized rapidly in the presence of oxygen when wet but more slowly when dry, are filtered off, washed with dry acetonitrile, and dried a t room temperature under vacuum. Anal. Calcd for CrBr2,2CH3CN: Cr, 7.69; Br, 54.38. Found: Cr, 7.7; Br, 54.5. Attempts to prepare CrBr2.2CH3CNby reducing a chromium(111) bromide acetonitrile solution6 with Zn-Hg resulted in a mixture of blue and white (presumably3 ZnBr2.2CH2CN) crystals. Diiodobis(acetonitrile)chromium(II).--A solution of chromium(II1) iodide in acetonitrilea is reduced by Zn-Hg to give a bright blue s o l ~ t i o n . ~ The hot solution is filtered, and the pale blue crystals which separate on cooling are filtered off and dried under vacuum. The acetonitrile solution and the solid complex are oxidized immediately in the presence of trace amounts of oxygen. Anal. Calcd for CrI2.2CH3CN: C, 12.38; H , 1.56. Found: C, 12.5; H, 1.6. Blue crystals, rapidly oxidized in air, are obtained by recrystallizing CrIz.BH20 from acetonitrile. They appear to have the approximate composition Cr12.2CH2CN.3H20and were not investigated further because of their extreme sensitivity to trace amounts of oxygen. Anal. Calcd. for Cr12.2CHaCN.3H20: C, 10.87; H , 2.74; Cr, 11.77; I , 57.42; N, 6.34. Found: C, 10.7; H , 2.6; Cr, 12.0; I, 59.5; N, 6.0. Attempted Isolation of Tetrahalo Chromium(11) Complexes.The addition of tetraalkylammonium chloride to acetonitrile and/ or ethanol solutions of CrClz.2CH3CN resulted in a solution (7) A. Naumann, Chem. Bet'., 47, 247 (1914). (8) R. D. Willett and R . E. Rundle, J . Chem. Phys., 40, 838 (1964). (9) J. P. Fackler, Jr., and D. G . Holah, I m v g . Ckenz., 4, 954 (1965).
Inorganic Chemistry from which only starting materials were recovered. With CrBrz. 2CH3CN, the addition of tetraalkylammonium bromide (in quantities we estimate to greatly exceed twice the molar quantity of CrBr2.2CH3CX)produced a pale yellow solution, suggestive of the formation of a new species (vzde znfm). But here a150 only starting materials were obtained on evaporation or cooling. With CrIz.2CH3CN, tetraalkylammonium (or arsonium) iodide addition produced some oxidation of the chromium(I1).
Results Dichlorobis(acetonitrile)chromium(II) is easily prepared in a pure state but decomposes in minutes in air.lO The bromide complex appears to be more reactive than the chloride, while CrIy2CH3CN is by far the most airsensitive chromium(I1) compound we have preparedl>Q,10 to date. As a result, i t was very difficult to study. While small amounts of water generally do not interfere with the coordinating ability of acetonitrile, especially when the latter is in a large excess,3 CrBr2. 2CH3CX must be prepared under anhydrous conditions, otherwise complexes containing varying amounts of coordinated water are produced. Recrystallization of CrIz.6H20from acetonitrile also gives a product containing both n ater and acetonitrile. Infrared spectra in the 2300 cm-l region show the acetonitrile to be coordinated through nitrogen as in other acetonitrile complexes.3-6 The chloride and bromide complexes have room temperature moments of 4.81 BM (xgullcor= 47.1 X cgs a t 22') and 4 83 BM (xgUnc"' = 32.7 X cgs a t 26'), respectively. The principal visible and near-infrared absorption spectra maxima for each of the complexes, with their solution molar extinctions, are recorded in Table I. Reflectance spectra of the chloride and bromide complexes are quite similar, both in the shape and position of the bands. In both complexes, the low-energy shoulder is clearly resolved into a maximum a t low temperatures. These spectra are 17ery similar to the spectra of CrC12.2py and CrBr2.2py, which have absorption maxima a t 15.050 and 10,800 cm-' and a t 14,300 and 10,300 cm-I. respectively.1° The reflectance spectrum of Cr12*2CH3CN,however, is significantly different from the spectra of the chloride and bromide. ,4t both room and low temperatures the lomenergy band appears as a broad maximum, in almost the same position as in the chloride and bromide, while the second band is shifted 3000 cm-I to higher energies. X-Ray powder patterns of CrC12.2CH3CN and CrBr2.2CH3CNshow that the two complexes are not isomorphous. Neither pattern resembles those of the two isomorphous. polymericlo CrX2.2py complexes. Solution spectra maxima of the complexes in acetonitrile m ith and without added halide ions are also recorded in Table I. CrC12.2CH3CKis only sparingly soluble in acetonitrile, but a spectrum was obtained with a saturated solution in 10-cm cells. The band maximum appears in almost the same position as in the reflectance spectrum; however, the band in the 10,000cm-I region is seen only as a slight asymmetry on the low-energy side of the main band. The solution spectrum of (lo) D. G . Holah and J. P. Fackler, Jr., Inovg. Chew., 4, 1112 (1965)
NOTES 431
VoL 5, No. 3, March 1966 TABLE I
ABSORPTION MAXIMA ( C M - ~ ) OF CrX2.2CHZCN COMPLEXES With tetraalkylammonium halided
CrC12.2CHaCN Reflectance 0.0016 M CHaCN 0.0282 M CzH50H CrBrz 2CH3CK Reflectance 0.0045 M CHaCiY CrI*.BCH&N Reflectance
13,3OOb 13,9OOc 13,200 (35)a 12,500 (34)
9,800 shb 9,9000 ~ 1 0 , 0 0 sh 0 -10,400 sh
13,200b 13,7OOc 11,800 ( -80)
9,600 shb 9,3OOc -10,000 sh
13,200 sh
10,100 ( ~ 5 0 ) 11,600 (-40)
10,000 (-230)
10,000b 16,200b 16,7OOc 10,4OOc 15,400 (3.6) 10,000 (-1) 0,279 M CHaCN b Room temperature. c Liquid nitrogen a Molar extinction coefficients, when available, are given in parentheses; sh, shoulder. Excess estimated to be much greater than twice the concentration of CrXz. Additional halide produces no further temperature. changes.
Discussion
I--" - 200
-150 Io -
I
= X
Figure 1.-Spectra of CrBrz. 2CHaCN: A, reflectance, 77'K; B, 0.0045 M in CHaCN solution; C, solution B plus large excess of ( CzH5)4NBr.
CrBr2.2CH3CNin acetonitrile differs considerably from the reflectance spectrum (see Figure 1). When excess tetraalkylammonium halide is added to the solvent, pronounced effects on the spectra of the complexes are observed. With CrC12.2CH3CN in acetonitrile containing (CzH5)4NC1, the solubility increases and the absorption maximum shifts to -10,100 crn-1 ( E -50). However, a shoulder remains at the same position as the band maximum before halide addition. Similar but less pronounced shifts are observed in ethanol. With CrBr2.2CH3CN, addition of halide (see Figure 1) shifts the band from the visible nearly completely into the infrared. The color of the solution changes accordingly from blue-green to pale yellow. A significant increase in the molar extinction also is observed. Addition of halide ions t o the iodide solution in acetonitrile gives some oxidation. Although no accurate measurements could be obtained, the 15,400 cm-l band appears to move to lovr7er energies with an increase in extinction coefficient.
A general discussion of the spectra of 3d4systems has been presented previously. Since the reflectance spectra of CrCI2.2CH3CN and CrBr22CH3CN are very similar to the spectral0 of CrXz.2py (X = C1, Br), similar assignments of the bands are suggested. If the low-energy band is the transition between the split components of the 6E, ground state, as is thought to be the case in the hydrate^,^ then the higher energy band, which contains the transitions to the split 5T2gstate, includes the transition with an energy corresponding to 1ODq. The shift to higher energy in the pyridine complex is consistent with the slightly higher position of pyridine relative t o acetonitrile in the spectrochemical series. Based on spectral evidence, the environment of the chromium(I1) in CrX2.2CH3CN(X = C1, Br) appears to be similar to that in CrX2.2py (X = C1, Br), i.e., distorted octahedral with four halides and two CH3CN groups surrounding the chromium(I1). The X-ray powder patterns of CrX2.2CH3CN(X = C1, Br) contribute little additional information to the structural problem except to indicate that some differences between the structures of the chloride and bromide must exist. The position of the main band in the reflectance spectrum of Cr12.2CH3CN a t -3000 cm-1 higher energy than the band in chloride and bromide complexes suggests a major structural change occurs with the iodide. The molar extinction coefficients at the band maxima in acetonitrile solution (Table I) also are significantly different. The low values observed (E -1-4) are consistent with the presence of an essentially centrosymmetric structure in solution. These molar extinctions are similar t o those for Cr2+(aq)where a molar extinction of 5.0 was obtainedg for the principal band. The reflectance and solution spectra of Cr12.2CH3CNsuggest that site symmetry of the chromium(I1) is similar in the solid state and in solution. This suggests the absence of large quantities of Cr(CH3CN)62+ ions in solution. The chromium(I1) halide-acetonitrile complexes were studied in the presence of excess tetraalkylammonium (or arsonium) halides in the hope of isolating
482
NOTES
solid tetrahedral chromium(1I) compounds. Gruen and McBethll presented spectral evidence for such a tetrahedral species from the spectrum of CrClz in a KC1-LiC1 eutectic. They assigned a band at 9800 &Tz absorption in T d symcm-l ( E -45) to the :E metry. Other bivalent first-row transition metal ions give typical tetrahedral spectra in this melt. In a CdS(s) crystal, the spectrumI2 and paramagnetic resonanceI3 have suggested that the chromium(I1) occupies a slightly distorted tetrahedral site. A spectral band near 5500 cm-‘ is reported.12 Thus a 4300 cm-1 separation exists between what presumably is the 5E + 5T2(Td)absorption of chromium(I1) in the two media studied. This apparent discrepancy could be due to the presence of a distorted tetrahedral arrangement (compare with the data on CuC14*- in ref 1) about the chromium(I1) in the eutectic. The spectrum14 of CrClz in anhydrous A1Cl3(1), where octahedral holes exist, shows a band slightly different in position from the band in the eutectic. Cnfortunately, intensities could not be measured.14 Anhydrous CrC12, which contains six-coordinate16 chromium(II), also shows a bandg at nearly the same frequency. The reflectance spectrum suggests the intensity ( E -1-10) of this band is not significantly different from thatg of other six-coordinate chromium(11) complexes, however. Unlike copper(I1) chloride,16 there is no evidence for MC142- formation in acetonitrile or ethanol on addition of excess chloride. With CrBrz.2CH3CN, however, spectral changes consistent with solvolysis17occur in acetonitrile (Figure I). Addition of bromide causes a shift of the principal visible absorption to lower energy and a fourfold increase in the extinction coefficient. Qualitatively similar changes occur when tetrahedral MX4*- complexes of the other bivalent transition metals are formedl8 in this solvent. While the extinction coefficient rules out the formation of a centrosymmetric six-coordinate CrBr64ion, the absorption appears at rather higher energies than predicted for a regular tetrahedral species and, in fact, appears near the energy of the CrCId2-band in the KCl-LiC1 eutectic.ll Thus one reasonably concludes (11) D. M. Gruen and R. L. McBeth, P w e A p p l . Chem., 6 , 23 (1963). (12) R. Pappalardo and R. E. Dietz, Phss. Rev., 125, 1188 (1961). (13) T. L. Estle, G. K. Walters, and M. DeWit, “Paramagnetic Resonance,” w. Low, Ed., Academic Press Inc., New York, N. Y . , 1963, p 144. (14) H. A. Dye and D. M. Gruen, Inovg. Chem., S, 836 (1964). ( 1 5 ) J. W. Tracy, N.W. Gregory, E. C. Lingafelter, J. D. Dunitz. H. C. Mez, R. E. Rundle, C. Scheringer, H. L. Yakel, Jr., and M. K. Wlkinson, Acta Cryst., 14, 927 (1961). (16) A further difference between copper(I1) and chromium(I1) is obvious from the acetonitrile complexes themselves. T h e chromium(I1) halides form complexes with two CHICN molecules while copper(I1) chloride and bromide form mono(acetonitri1e) complexes. Attempts t o prepare’ CuClv 2CHaCN were not successful. (17) Solvolysis of some transition metal halides in nonaqueous solvents has been discussed previously. For example, the articles by S. Buffagni and T. M .D u m , J . Chem. Soc., 5106 (19611, and D. A. Fine, J . A m . Chem. Soc., 84, 1139 (1962), consider the various complexes formed with cobalt(I1) halides in a number of solvents, i . e . , [L’rXala-, [MXaSl-, [MXzSz], etc., where X = C1, Br, I and S = solvent. Equilibria, leading to the formation of similar complexes, are thought to exist in the C r B r r C H C S solution. (IS) PIT. S. Gill and R. S. Kyholm, J . Chem. Soc., 3997 (1969). These authors also stated that a later paper would consider tetrahedral C r ( I I ) , h u t such a paper has not appeared.
I.lzorgnnic Chernistry that a distorted tetrahedral anion predominates in the solution. As indicated previously, all attempts to isolate a tetrahedral CrX42- complex failed. Unfavorable radius ratioslg may be responsible in part for this since chromium(1I) is the largest bivalent, first-row transition metal ion which gives any evidence a t all for tetrahedral anion formation. We are currently attempting to correct these unfavorable size effects by preparing complexes of the type L2CrX2,n-here L is sterically large. Acknowledgments.-The support of the Kational Science Foundation, GP-4253, and the donors of the Petroleum Research Fund is gratefully appreciated. (19) L. Pauling, “The Nature of the Chemical Bond,” 2nd ed, Cornel1 University Press, Ithaca, N. P., Chapter X.
COXTRIBUTION PROM THE DEPARTMENT OF CHEMISTRY, THE
UXIVERSITY OF MISSOURI, COLUMBIA, ~~~ISSOURI
Polarographic Investigation of the Allyl Alcohol Complex of Copper(1) in Aqueous Solution BY STAXLEY E. MANAHAX
Receiaed Sefltenzbev 8,1965
An investigation of the solubility of cuprous chloride in aqueous solutions of allyl alcohol has shown that copper(1) ion forms a 1: 1 complex with allyl a1cohol.l A stability constant of lo4.’ was determined for the species Cu(a1c) +. The investigation was complicated, however, by the formation of the chloride complexes CuC12- and Cu(a1c)Cl. Using polarography we have determined the formation constant of the allyl alcohol complex of copper(1) in aqueous media 0.1 M in SaC104. The concentration of allyl alcohol ranged from 1.00 X to 1.00M . Half-wave potential values were obtained for both the Cu(I),Cu(Hg) couple and the Cu(II),Cu(I) couple a t both the dropping mercury electrode and the dropping copper amalgam electrode. In the former case copper(11) was added to the solution as the perchlorate and in the latter case copper was not present in solution. The values of Ea14- El,, for all the polarographic waves fell within the range of 0.056-0.060v, indicating reversibility of the electrode reactions. The half-wave potentials measured a t the mercury electrode were in close agreement with the half-wave potentials obtained a t the amalgam electrode. The polarographic reduction of Cu2+ to Cu(Hg) at the dropping mercury electrode (or oxidation of Cu(Hg) to Cu2+ a t the dropping copper amalgam electrode) in aqueous noncomplexing media proceeds via a one-step, two-electron process. The addition of (1) R. M. Keefer, L. J. Andrew, and R . E. Kepner, J . Am. Chem. SOC.,71, 3906 (1949). (2) J . Tomes, Collectioiz Czech, Chem. Commun., 9, 12, XI, 150 (1937).
